broadfield digital founder | architecting ai for brands
Joined June 2025
- Tweets4.4K
- Following285
- Followers50.7K
- Likes7.6K
Pinned Tweet
Absolutamente recomiendo usar Cache Control en Claude Code
Probablemente uno de los Mods que te ahorrará la mayor cantidad de tokens y ayudará a mantener tus sesiones activas por mucho más tiempo
Añade esta barra para la caché de 5 minutos de Claude y envía notificaciones para ayudar a mantener la caché activa
Comando de instalación:
npx claude-code-templates@latest --mod observability/prompt-cache-control
Stefan Friedman retweeted
me calling support last year: "please hold while I transfer you"
me calling an agent on Phonic-1: it already fixed it while I was still explaining the problem
10 person team btw
Introducing Phonic-1, the most advanced speech-to-speech model for agentic tasks — from @phonic_co, the speech-to-speech model company.
We’ve had a breakout year and raised a $27.5M Series A led by @firstround with participation from @Lux_Capital, @felicis, and Definition.
Stefan Friedman retweeted
EVERY GROUP CHAT HAS THAT ONE TRIP WITH 37 MESSAGES AND ZERO FLIGHTS BOOKED
Someone sends a meme and the whole plan disappears.
In Wabi every topic gets its own thread, and the AI in that thread messages you first. So the trip lives in its own chat that keeps bugging everyone about the flights until someone books them.
I'm putting the group in one tonight. Taking bets on who leaves first.
Today, I’m excited to share that we’ve raised $55M in Series A funding at a $1.25B valuation, led by @IVP, with support from @a16z, @ForerunnerVC, and an incredible group of founders and investors including @Benioff, @tobi, @dickc, and @marissamayer.
Everyone has an operating system for their phone. We’re building one for your life.
Right now, life is scattered across apps, group chats, calendars, and tabs. You’re the one connecting the dots and keeping everyone on the same page.
@wabi is a personal AI that helps you get your whole life organized. It brings your people, context, and apps together so you can coordinate plans, share the mental load, discover new things, and show up for each other.
Because AI shouldn’t just help us get things done. It should help us do more of what we love, with the people we love.
Join us 👇
Stefan Friedman retweeted
one sentence in, full investor deck out
I don't use Muse anymore. Robyn built the agent, pulled my context and handed me something I'd actually send
🤝 Paid partnership
Meet Robyn, the personal agentic platform built around your life and work.
- Tell her what you need in one sentence. No setup, ever.
- Spin up a fleet of agents preloaded with your context that work while you sleep.
- Access it on the web or just text it.
- 20hrs/week saved per user.
I’m Jenny, and I spent 3 years working with Nobel Laureate Eric Kandel, who mapped out memory in the human brain.
We built all of that into Robyn, making it the first AI capable of taking intelligent action.
Invite only access opens today. We’re giving out 1000 invite codes. Link in first comment. Comment "Robyn" for 100 free credits.
Stefan Friedman retweeted
OPENROUTER HAS A REAL PROBLEM NOW...
RouterPlus just launched and it doesn't pick models off the price list
you set a quality bar, it scores every model on your own evals, and each type of task goes to the cheapest one that clears it
so your support triage could end up on qwen while code review stays on gpt
and with your own keys the gateway fee is 0 at any volume. openrouter takes 5% once you pass $25k a month
did not expect a gateway launch to be this interesting
AI teams overpay for inference twice: once for models bigger than they need, and again for idle capacity they don’t use. @useRouterPlus is the extended inference team for fast-growing AI companies. We test models and providers on your traffic and get providers to compete for your work.
Make your inference better. Go to routerplus.com
Stefan Friedman retweeted
one sentence in, full investor deck out I don't use Muse anymore.
Robyn built the agent, pulled my context and handed me something I'd actually send
Introducing Cartesia Managed Agents: our end-to-end voice agent platform for enterprise.
#1 on @ArtificialAnlys Speech Agent Arena and @ServiceNow EVA-Bench, built by co-designing our voice models and harness as one system.
Check it out → cartesia.ai/blog/managed-age…
Stefan Friedman retweeted
not another AI that spits out a new design every time you ask for a change
what canva's brush actually does:
1. reads how every element of your design fits together
2. plans the edit
3. restyles it, adapts it to a new format or updates every page
4. runs under the hood of canva AI, built by canva itself
second in-house model in two weeks btw
🤝 Paid partnership
Stefan Friedman retweeted
the guy who spent his weekend regenerating a video just to get it in 4K watching this
Introducing P-Video-Upscale! 🎬
Fast video upscaling to 720p, 1080p, 2K, or 4K, while preserving the original aspect ratio and audio. Need smoother motion? Increase the frame rate to up to 60 fps.
And for launch, it's 50% off for the first two weeks, until November 12 at 18:00 CET!
Here's what you need to know:
• 𝗣𝗿𝗶𝗰𝗶𝗻𝗴: Starting from $0.01/s at 720p
• 𝗦𝗽𝗲𝗲𝗱: ~1.7s per second of output video, depending on resolution
• 𝗙𝗿𝗮𝗺𝗲 𝗿𝗮𝘁𝗲: Keep the original or interpolate up to 60 fps
• 𝗧𝘂𝗿𝗯𝗼 𝗺𝗼𝗱𝗲: Faster processing with a slight quality trade-off
No need to regenerate a video just to get it in higher resolution.
Available via our inference partners @scenario_gg, @replicate, @WiroAI, @TellersAI, @eachlabs, @wavespeed_ai, Sociaro, and @runware.
Benchmarks by our benchmark partners @datapointai and @RapidataAI are coming soon.
📚 Model docs: docs.api.pruna.ai/guides/mod…
🧩 Try the model through the API or Playground: dashboard.pruna.ai/login
Stefan Friedman retweeted
the real cost in building design isn't the first sketch
it's the client moving one staircase in week eight, and someone spending days updating every drawing by hand
snaptrude agent works on projects already underway: you mark up the change, it tests the options, shows the impact and updates the full drawing set
that's where the hours actually go
try it: snaptrude.com
🤝 Paid partnership
Today we're launching @snaptrude Agent, the most advanced agent for building design.
It does real design work with you, from a site to a construction-ready LOD 400 model with structure, MEP and drawings. Then can go back to your existing software with your families intact.
About a year ago we bet the company on AI. Honestly the insight came from software engineering. It wasn't long ago that autocomplete felt like a big deal, and now engineers direct agents that write a lot of the code. I think the same shift is coming to how we design buildings, and it's already started.
Most designers I meet aren't short on skill, they're short on time. Take the call every design team dreads, the client cuts the budget by 30%. Today that's a bad month for everyone. With the agent you're the director. You decide where the cost comes out, and it updates the layout, coordinates the structure and MEP, and redoes the drawings. So when design exploration gets this cheap, we go from 3 options to 300, and more options means better buildings.
Since August our users have gone through more than 140 billion tokens, and firms like Jacobs, Studio+, Kajima, VMDO, Dekker and hundreds of others use Snaptrude today. In our early access, more than 2,000 designers have run over 25,000 tasks through the agent since September.
We also built a benchmark for this, AECBench, because in AEC an almost correct answer is still wrong. It gives AI the same brief an architect would get, has its design in a real BIM model, and scores the result against the brief and against how architects review design. We're publishing the first results today, link below.
Thank you @prashanthp at @Accel and Shub and the team at @foundamental for backing us when this was still a bet, and the Snaptrude team, who built every bit of this. Really proud of you all.
Thanks to our investors, @scottbelsky, @andreasklinger, @thilokonzok, @immad, @CindyXBi, @DanilaVladi, @jmwass, rahulchaudhary, @adamecourt, Zach Ginsburg, Alex duff, Paul grossinger, really appreciate all the help along the way.
Stefan Friedman retweeted
how most companies run AI agents right now:
→ agent uses the access of whoever set it up
→ nobody tracks what it touched
→ security says no to everything
with Mintive:
→ each agent only gets what its job needs
→ every action is logged, person to agent to action
→ one policy across Claude, ChatGPT, Copilot, Cursor
congrats on the $11M 👏
🤝 Paid partnership
Stefan Friedman retweeted
Every Al game launch I've seen spikes and then breaks. RouterPlus lets you reserve capacity before the spike and keeps a backup provider warm.
AI teams overpay for inference twice: once for models bigger than they need, and again for idle capacity they don’t use. @useRouterPlus is the extended inference team for fast-growing AI companies. We test models and providers on your traffic and get providers to compete for your work.
Make your inference better. Go to routerplus.com
Stefan Friedman retweeted
asked Robyn for every company that raised $50M+ this month
what came back:
1. sorted by sector
2. lead investors
3. what each one is building
a week of associate work. Instinct is done for me
🤝 Paid partnership
Meet Robyn, the personal agentic platform built around your life and work.
- Tell her what you need in one sentence. No setup, ever.
- Spin up a fleet of agents preloaded with your context that work while you sleep.
- Access it on the web or just text it.
- 20hrs/week saved per user.
I’m Jenny, and I spent 3 years working with Nobel Laureate Eric Kandel, who mapped out memory in the human brain.
We built all of that into Robyn, making it the first AI capable of taking intelligent action.
Invite only access opens today. We’re giving out 1000 invite codes. Link in first comment. Comment "Robyn" for 100 free credits.
Stefan Friedman retweeted
> be the Stanford researchers who invented state space models
> everyone builds on transformers
> bet on a different architecture anyway
> ship Sonic, #1 user preference on Artificial Analysis
> ship Ink 2, most accurate streaming STT
> today: Managed Agents, #1 on ServiceNow's EVA-Bench
> prompt in, phone call out
Introducing Cartesia Managed Agents: our end-to-end voice agent platform for enterprise.
#1 on @ArtificialAnlys Speech Agent Arena and @ServiceNow EVA-Bench, built by co-designing our voice models and harness as one system.
Check it out → cartesia.ai/blog/managed-age…
Stefan Friedman retweeted
genuine question for anyone building with ai video
how much were you paying per clip before today
just ran the same seedance clip through sunra and i don't want to say my old number out loud
Stefan Friedman retweeted
most AI for buildings stops at the massing, where mistakes are cheap
snaptrude agent goes into the details:
• give it a manufacturer's spec sheet and it models that exact glass facade, down to the anchors and bolts
• it adds the structure, plumbing and wiring
• it drafts the drawings builders work from
and the architect stays in charge the whole way
try it yourself: snaptrude.com
🤝 Paid partnership
Today we're launching @snaptrude Agent, the most advanced agent for building design.
It does real design work with you, from a site to a construction-ready LOD 400 model with structure, MEP and drawings. Then can go back to your existing software with your families intact.
About a year ago we bet the company on AI. Honestly the insight came from software engineering. It wasn't long ago that autocomplete felt like a big deal, and now engineers direct agents that write a lot of the code. I think the same shift is coming to how we design buildings, and it's already started.
Most designers I meet aren't short on skill, they're short on time. Take the call every design team dreads, the client cuts the budget by 30%. Today that's a bad month for everyone. With the agent you're the director. You decide where the cost comes out, and it updates the layout, coordinates the structure and MEP, and redoes the drawings. So when design exploration gets this cheap, we go from 3 options to 300, and more options means better buildings.
Since August our users have gone through more than 140 billion tokens, and firms like Jacobs, Studio+, Kajima, VMDO, Dekker and hundreds of others use Snaptrude today. In our early access, more than 2,000 designers have run over 25,000 tasks through the agent since September.
We also built a benchmark for this, AECBench, because in AEC an almost correct answer is still wrong. It gives AI the same brief an architect would get, has its design in a real BIM model, and scores the result against the brief and against how architects review design. We're publishing the first results today, link below.
Thank you @prashanthp at @Accel and Shub and the team at @foundamental for backing us when this was still a bet, and the Snaptrude team, who built every bit of this. Really proud of you all.
Thanks to our investors, @scottbelsky, @andreasklinger, @thilokonzok, @immad, @CindyXBi, @DanilaVladi, @jmwass, rahulchaudhary, @adamecourt, Zach Ginsburg, Alex duff, Paul grossinger, really appreciate all the help along the way.
Stefan Friedman retweeted
My mom called a delivery company last week, switched to Portuguese halfway through because she got nervous, and the bot just hung up on her lol
Phonic-1 switches between 50+ languages mid call, same voice. and it's $0.06/min
someone send this to every support team pls
Introducing Phonic-1, the most advanced speech-to-speech model for agentic tasks — from @phonic_co, the speech-to-speech model company.
We’ve had a breakout year and raised a $27.5M Series A led by @firstround with participation from @Lux_Capital, @felicis, and Definition.
Stefan Friedman retweeted
fastest way to spot an AI ad is still the text
there's always a "LIMTED OFER" sticker somewhere and a font no brand on earth uses
krea ad studio writes the headline in the brand's own font, so the hand wash one at the end reads like a real shop made it. curious if it holds up on smaller products
first few ads are free, link below
🤝 Paid partnership
fastest way to spot an AI ad is still the text
there's always a "LIMTED OFER" sticker somewhere and a font no brand on earth uses
krea ad studio writes the headline in the brand's own font, so the hand wash one at the end reads like a real shop made it. curious if it holds up on smaller products
first few ads are free, link below
🤝 Paid partnership
Stefan Friedman retweeted
a glowing "70% OFF" sticker in the corner is how i know someone made their ad in chatgpt
this hoodie went through krea ad studio instead
>paste the store link
>it finds the hoodie by itself
>ad comes back with the brand's real font on it
no idea yet how it does on meta next to what we run now. first few ads are free so that's this week's test
link in the reply
🤝 Paid partnership · 🤖 Made with AI